It seems that we have a discussion where some (without saying so) approach the discussion from a low volume standpoint and some (without saying so) approach the discussion from a high volume standpoint.

When I were to produce 10 units, the deciding factors were a) tools already in the shop and b) knowledge of the processor. Producing 10 units it does not matter if the parts cost go up by $50 if you can save a week of development time.

When I were to produce 500.000 units we spent 2 months to make the right decision regarding hardware and then another month procuring tools and familiarizing those of us not programming application software in C with the intricacies of that hardware.

Again, the PIC :-) of processor etc depends 90% on the volume to be produced.

A side remark to a subject touched in a few of the postings: I will NOT use a processor for which the only toolset is the manufacturers. Chip manufactureres make excellent chips, Compiler manufactureres make excellent compilers, Editor manufacturers make excellent editors. Thus I want to buy from the right source. Of course, If I were a hobbyist, this would be different.

Have fun,

Erik